Joint Collaboration Team-Video Coding (JCTVC), which is a joint standardizing organization of ITU-T and ISO/IEC, has standardized an image encoding scheme referred to as high efficiency video coding (HEVC) for the purpose of further improving encoding efficiency than H.264/AVC (for example, see NPL 1 below). HEVC provides not only encoding of a single layer but also scalable video coding. The scalable video coding technology for HEVC is also referred to as scalable HEVC (SHVC) (for example, see NPL 2 below).
In general, the scalable video coding refers to a technology for hierarchically encoding layers that transmit rough image signals and layers that transmit fine image signals. The typical attributes hierarchized in the scalable video coding of the related art are 3 types of attributes: spatial resolution (spatial scalability), a frame rate (temporal scalability), and a signal-to-noise (SN) ratio (SNR scalability). Further, NPL 3 proposes color gamut scalability in which a color gamut, which is a kind of pixel attribute, is hierarchized. A bit depth can also be hierarchized along with the color gamut. NPL 4 proposes a technology for hierarchizing a luminance dynamic range which is also a kind of pixel attribute.
According to a scheme proposed in NPL 3, corresponding pixel vectors in the color gamut (for example, ITU-R BT.2020) of an enhancement layer are predicted from pixel vectors (Y, U, and V) in the color gamut (for example, ITU-R BT.709) of a base layer using a 3-dimensional lookup table. The lookup table is expected to be adopted in the name of a color mapping table in the standard specification of SHVC. In the color mapping table, a color space of a base layer is split into several cuboid partitions. In the color mapping table, the plurality of respective partitions formed through the splitting are mapped to corresponding partial spaces of the color space of the enhancement layer. One partial space can be specified by defining four vertexes. According to the scheme proposed in NPL 3, specific predicted pixel values are decided through tetrahedral interpolation based on the pixel values of the four vertexes.